AWS launches Agentic Shopping Assistant, powers retailers' AI agents in 60 days

Amazon Web Services (AWS) has introduced the AWS Agentic Shopping Assistant (ASA), a new service that lets retailers build and deploy their own AI-powered shopping agents. The tool is built on the same technology that powered Amazon's Alexa for Shopping, giving retailers access to Amazon's proven conversational retail AI without having to develop it from scratch. According to AWS, retailers can launch a customized agent in approximately 60 days, significantly reducing time-to-market. The ASA is designed to handle tasks like product recommendations, order tracking, and answering customer queries, operating as a specialized agent within retail environments.

Kate Spade, a luxury fashion brand, is among the first retailers to adopt the ASA. Their AI agent runs on Anthropic's Haiku 4.5 model, accessed through Amazon Bedrock — AWS's managed service for foundation models. This combination allows Kate Spade to deliver a tailored, conversational shopping experience while maintaining control over their brand voice and data. The launch signals AWS's broader strategy to commercialize its internal AI capabilities and compete with other retail AI platforms. By offering a pre-built agentic framework, AWS is targeting retailers who want to quickly implement AI shopping assistants without heavy investment in proprietary development.